Cancer genetics has revealed that p53 and LKB1/STK11 are the most commonly mutated tumor suppressors in sporadic human non-small cell lung cancers (NSCLC), the leading source of annual cancer deaths in the U.S. LKB1/STK11 encodes a Ser/Thr protein kinase that directly phosphorylates the activation loop of the AMP-activated protein kinase (AMPK) as well as 12 poorly understood related kinases in the AMPK family. AMPK is a master regulator of cellular and organismal metabolism that acts as a sensor of cellular energy, arresting cell growth and reprogramming metabolism when ATP levels are low. Over the past 5 years, a number of labs including ours have decoded substrates of AMPK and related kinases that mediate downstream effects on growth and metabolism and may relate to the tumor suppressor activity of LKB1, including AMPK phosphorylation of core components in the mammalian target of rapamycin (mTOR) and autophagy pathways. In addition, the front-line type 2 diabetes drug metformin has been shown to regulate cell growth in an AMPK- and mTOR-dependent manner in some settings, suggesting it may serve as a potential anti-cancer agent. Despite these direct connections between AMPK and growth regulators, there is a great deal of overlap between the downstream functions and effectors of AMPK and its 12 related kinases, so it remains unclear which of these 14 kinases that LKB1 directly activates are the most critical for mediating its tumor suppressor function. Moreover, accumulating evidence suggests that in many settings the ability of AMPK to restore metabolic homeostasis under glucose or oxygen-poor conditions may promote survival of cancer cells. Thus, the role of AMPK in tumorigenesis may be very context dependent, and a different AMPK related kinase may be more important for the ability of LKB1 to suppress NSCLC. Finally, while epidemiological data and mouse xenograft and tobacco carcinogen models support a beneficial effect of metformin, this has not been examined in a genetically engineered mouse model of a human cancer in a manner that allows one to distinguish genotype-specific therapeutic effects. Moreover, metformin and its more potent analog phenformin are mitochondrial inhibitors that affect pathways outside of AMPK, and may selectively allow for the killing of LKB1-deficient tumors as is observed in cell culture models. The specific aims are to 1) define which of the 14 AMPK family kinases are essential for the ability of LKB1 to suppression tumorigenesis in a NSCLC xenograft model; 2) genetically define the role of AMPKa1 or AMPKa2 and related family kinases in a genetic engineered mouse model of NSCLC; and 3) examine the therapeutic efficacy and genotype selectivity of AMPK-activating biguanide compounds metformin and phenformin in multiple genetic engineered mouse models of NSCLC.